CN1115828C - Pre-session pre-synchronized framing method for real-time services in ATM networks - Google Patents
Pre-session pre-synchronized framing method for real-time services in ATM networks Download PDFInfo
- Publication number
- CN1115828C CN1115828C CN97120230A CN97120230A CN1115828C CN 1115828 C CN1115828 C CN 1115828C CN 97120230 A CN97120230 A CN 97120230A CN 97120230 A CN97120230 A CN 97120230A CN 1115828 C CN1115828 C CN 1115828C
- Authority
- CN
- China
- Prior art keywords
- node
- time
- unit
- delay
- frame
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Expired - Fee Related
Links
- 238000000034 method Methods 0.000 title claims abstract description 45
- 230000001360 synchronised effect Effects 0.000 title claims abstract description 8
- 238000009432 framing Methods 0.000 title description 9
- 238000011144 upstream manufacturing Methods 0.000 claims description 36
- 230000008878 coupling Effects 0.000 claims description 16
- 238000010168 coupling process Methods 0.000 claims description 16
- 238000005859 coupling reaction Methods 0.000 claims description 16
- 230000001934 delay Effects 0.000 abstract description 5
- 238000005516 engineering process Methods 0.000 description 6
- 238000004891 communication Methods 0.000 description 5
- 230000005540 biological transmission Effects 0.000 description 2
- 230000002123 temporal effect Effects 0.000 description 2
- 101100311260 Caenorhabditis elegans sti-1 gene Proteins 0.000 description 1
- 241001522296 Erithacus rubecula Species 0.000 description 1
- 238000009825 accumulation Methods 0.000 description 1
- 239000012141 concentrate Substances 0.000 description 1
- 239000008141 laxative Substances 0.000 description 1
- 230000002475 laxative effect Effects 0.000 description 1
- 230000014759 maintenance of location Effects 0.000 description 1
- 230000002265 prevention Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L49/00—Packet switching elements
- H04L49/30—Peripheral units, e.g. input or output ports
- H04L49/3081—ATM peripheral units, e.g. policing, insertion or extraction
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04Q—SELECTING
- H04Q11/00—Selecting arrangements for multiplex systems
- H04Q11/04—Selecting arrangements for multiplex systems for time-division multiplexing
- H04Q11/0428—Integrated services digital network, i.e. systems for transmission of different types of digitised signals, e.g. speech, data, telecentral, television signals
- H04Q11/0478—Provisions for broadband connections
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L12/00—Data switching networks
- H04L12/54—Store-and-forward switching systems
- H04L12/56—Packet switching systems
- H04L12/5601—Transfer mode dependent, e.g. ATM
- H04L2012/5638—Services, e.g. multimedia, GOS, QOS
- H04L2012/5646—Cell characteristics, e.g. loss, delay, jitter, sequence integrity
- H04L2012/5649—Cell delay or jitter
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L12/00—Data switching networks
- H04L12/54—Store-and-forward switching systems
- H04L12/56—Packet switching systems
- H04L12/5601—Transfer mode dependent, e.g. ATM
- H04L2012/5678—Traffic aspects, e.g. arbitration, load balancing, smoothing, buffer management
- H04L2012/5679—Arbitration or scheduling
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Data Exchanges In Wide-Area Networks (AREA)
Abstract
A method for transmitting jitter information in a network, wherein a connection for a call is established from a source node to a destination node through a plurality of nodes and a frame counter is located at each of the nodes, detects propagation delays and corresponding mismatch delays between neighboring node pairs, delay bounds at each of the nodes, and a frame size of the connection. Then the frame counters are synchronized and activated, based on the propagation delays and the corresponding mismatch delays between the neighboring node pairs, the delay bounds at a corresponding node, and the frame size of the connection, and a cell is serviced from the source node to the destination node through each node based on the value of the corresponding frame counter.
Description
Technical field
The present invention relates to a kind of being used for by utilizing the be ranked method of unit of ATM net of plenary session words phase frame; And, more specifically, relate to and a kind ofly can find out wobble information is not sent to this wobble information another node from a node with control shake thus method.
Background technology
B-ISDN (broadband integrated services digital network) is a kind of digital network, and this Netcom crosses that the user that will concentrate or disperse transmits according to a kind of broadband with server and switching technology couples together and provides such as the various services of serving in real time continuously and intensive data is served.In this B-ISDN process of realization, adopt ATM (asynchronous transfer mode) technology that information is transmitted by this User Network Interface, wherein this atm technology is a kind of packet-based asynchronous time division multi-plexing, and can utilize the advantage of legacy circuit-switched and packet switching to handle various services.
ATM net in the future need be provided at quantitatively and have huge many difference services of very different transmission characteristics qualitatively.Required service must be satisfied the strict demand aspect throughput, time-delay, shake and loss late; And because service has in real time become the main service of broadband networks, will be particularly harsh for the demand of time-delay and shake, because if do not transmit real time information in a time limit, this real time information will be considered to lose.
The relevant many researchs that can support the queuing art of the real-time communication service in the fast packet switching network have been reported.They can be divided into work conservative and inoperative conservative.In network with work reservation service principle, the offered load at node place fluctuation formerly may make the undesired and instant transmission rate at a certain node place that causes of the traffic pattern of connection increase, even satisfied the transfer rate restriction at this Web portal place of user's appointment when this connection.The surge of the traffic in connects along each saltus step accumulation from the connecting path of destination, a source to, needs more resources to be used for downstream node usually thus.
In order to correct this problem, advised that several inoperatives keep the packet service principle so that the packet jitter of standard at the intermediate node place increases sharply with the prevention traffic carrying capacity.Layering circle Rabin and stopping-transmission technology (Hierarch-ical round robin and stop-and-go techniqnes) is a typical example in them again; Time shaft is divided into permanent long frame and the frame of the disjunction of some is distributed to each connection.In this framing method, only the time frame that is assigned with is allowed packet service, and thereby can place restrictions on this delay jitter.Yet the method has bandwidth granularity and the time-delay coupled problem between placing restrictions in essence, thereby causes the application of a low bandwidth.A kind of shake expire the earliest and a kind of static precedence technique (a jitter earliest-due-date and a rate-controlledstatic priority techniques) of controlling transfer rate according to each grouping control packet jitter, be used to be sent to next node but need to calculate in guiding time of each grouping at each node place and it is marked at head.
With reference to figure 1, traditional shake technology that expires the earliest has been described.VT
I-1, j kBe that k unit arrives the virtual time of advent that the node (i-1) that connects j is located; DL
I-1, j kBe connect j in the deadline of k the unit that node (i-1) is located, be necessary for this unit service before between at this moment; STi-1, j be when connect j in k unit that node (i-1) is located by active service service time of this unit during in the next node (being node i) of node (i-1); And d
I-1, jBe that the time-delay of locating at node (i-1) that connects j is placed restrictions on.Further, ι i is the propagation delay between node (i-1) and the node i; AT
I, j kIt is the actual time of arrival that connects j in k the unit at node i place; HT
I, j kBe the retention time, k unit is temporarily stored at a dither controller during this period; DL
I, j kBe connect j in the deadline of k the unit at node i place, at this moment between before must serve this unit; ST
I, j kBe when connect j in k the unit at node i place by active service service time of this unit during in next node; And di, j is that the time-delay at the node i place that connects j is placed restrictions on.K unit node (i-1) locate serviced after, it is by with its deadline and service time, i.e. (DL
I-1, j k-ST
I-1, j k) between difference mark.K unit be made to be fit to be ranked before, the dither controller in the porch of node i keeps this section period HTi with k unit, j, and arrive virtual time of advent of the VT of the node i that connects j at k unit
I-1, j kThis k unit is sent to a scheduler.
Yet existing inoperative keeps the packet service principle with its current form and be not suitable for the application of atm network, because they are by utilizing framing method or controlling delay jitter by using packet header to transmit wobble information.The former is because its coupled problem in essence can not satisfy the various characteristics requirement that connects completely effectively, and the latter then requires a head space to send this time mark, and this almost is unallowed in the ATM net.
Summary of the invention
Therefore, a main purpose of the present invention provides a kind of method that can find wobble information this wobble information not to be sent to another node from a node to control shake thus.
According to an aspect of the present invention, a kind of method of the network unit that is used for being ranked is provided, wherein set up a connection that is used to call out by a plurality of nodes to destination node from source node, and one frame counter is set at each node place, this method comprise the following steps: (a) detect adjacent node between propagation delay and corresponding non-coupling time-delay, time-delay at each node place is placed restrictions on, and the frame sign that should connect, wherein each adjacent node points to the downstream node of destination node to comprising a upstream node and that points to source node, and this corresponding non-coupling time-delay is what to be caused by the non-coupling in the boundary of time slot between a upstream node and the downstream node; (b) according to adjacent node between propagation delay and corresponding non-coupling time-delay, the frame sign of placing restrictions on and should be connected in the time-delay at each node place, Active Frame counter also synchronously, each frame counter is positioned at each node place; And, transmit this unit from source node to destination node (c) according to the value of respective frame counter.
According to another aspect of the present invention, a kind of method is provided, be used for being lined up to Section Point from the first node of a network in a unit, this first and second node has one first and one second counter respectively, this method may further comprise the steps: (a) place restrictions on all counters of presynchronization according to the time-delay of propagation delay between the node and first node, the value of each counter begins to count down from the value by a presynchronization program setting and deducts 1, up to 1, be reset to M then, and begin to count down from M and deduct 1, up to 1, M is one greater than 1 integer, and the value that is counted corresponding to P is placed restrictions in this time-delay, P is equal to or less than M, wherein by the presynchronization process presynchronization of all counters is carried out as follows: the value sum of second counter when count value of first counter is same as (M-P) with this arrival Section Point place, unit in this unit leaves first node; (b) when this unit arrives first node, keep this unit to count down to 1 and be reset and be M up to first counter at this; And (c) any sends to Section Point with this unit from first node before taking place in two kinds of situations below in the count value of first counter: if P less than M, then this count value reaches (M-P); If P equals M, then this value is reset to M.
According to a further aspect of the invention, provide a kind of by the be ranked method of a plurality of unit of a calling of the connection that constitutes by a plurality of nodes, wherein these nodes comprise source node and destination node, each node of this connection has a frame counter, this method may further comprise the steps: the size with a frame is placed restrictions in a time-delay of (a) determining each node of this calling, wherein this calling is divided into a plurality of frames at each node place, each frame comprises M time slot, M is a positive integer, and each time slot is a default time interval; (b) determine a propagation delay between the two right nodes of each adjacent node, each adjacent node is to comprising a upstream node that points to source node and a downstream node that points to destination node; (c) place restrictions on all frame counter presynchronization according to the time-delay of the right upstream node of the right propagation delay of each adjacent node and each adjacent node to this node, the time-delay of this upstream node is placed restrictions on corresponding to P time slot, P one is not more than the positive integer of M, the count down time slot of each frame of this node of the frame counter of one node, wherein the right counter of each adjacent node by presynchronization in case the count value of this upstream node that a unit is left equal (M-P) and this unit arrive this downstream node the time this downstream node the count value sum; (d) when each unit arrives the right upstream node of the adjacent node that comprises source node, keep each unit to arrive the end of a frame up to this unit; When (e) count value of the upstream node in step (d) satisfies one of following two conditions each unit is sent to its corresponding downstream node from this upstream node: if P less than M, then this count value reaches (M-P); If P equals M, then reseting this count value is M; And (f) for each remaining adjacent node node that to repeating said steps (d) and (e) achieves the goal up to each unit.
Description of drawings
Above and other purpose of the present invention and characteristics will become apparent by the description to the following preferred embodiment that provides with accompanying drawing.
Fig. 1 illustrates traditional shake principle that expires the earliest;
Fig. 2 has shown the exemplary plot of the network that a plurality of nodes are formed;
Fig. 3 has shown the system layout of a node of PSPS service principle;
Fig. 4 provides frame counter synchronous sequential chart;
Fig. 5 represents first embodiment when the presynchronization of placing restrictions on the frame counter when equaling a frame sign in the time-delay at node place;
Fig. 6 has described second embodiment of the presynchronization of the frame counter when time-delay is placed restrictions on less than frame sign;
Fig. 7 has described the 3rd embodiment of the presynchronization of the frame counter when time-delay is placed restrictions on greater than frame sign; And
Fig. 8 provides the 4th embodiment of the presynchronization of the frame counter that is used for various propagation delays.
Embodiment
In a communication pattern of the present invention, a network is made up of one group of node that links up by one group of link, as shown in Figure 2.Each one way link allows two nodes by a bandwidth communication.One grouping that connects or unit are in the path that is called as this grouping from its source node to its destination node this group link that passes in service.Example among Fig. 2 has 2 connections; One connects J1, and the path of being followed is P1={AC, CD, DE}; Another connection J2, the road of being followed is P2={BC, CD, DF}.For link CD, using this group connection of link CD is { J1, J2}.
According to the present invention, designed a kind of new queuing policy that is suitable for real-time communication service, this strategy provides to have and guarantees that data volume communicates by letter with almost constant not stopping up of end-to-end time-delay.Queuing policy of the present invention has used a kind of plenary session words phase presynchronization (PSPS) framing strategy.During at the frame counter at each node place one each frame that connects being defined as each session, and the presynchronization of earthing, the border frame counter of this frame is connecting the establishment stage alignment.These a little synchronous frame counters can be controlled delay jitter effectively, and do not need to transmit the head space of this wobble information.Therefore, this PSPS framing strategy can be applied directly to the ATM net, and a kind of attractive scheme that critical temporal information transmits that solves is provided, and these critical temporal informations are the real time communication signals such as sound, video and circuit simulation in the ATM net.
Fig. 3 has shown the system layout of a node of PSPS service strategy.The system 30 of one node i comprises 32 to 34 and EDD of controller (expiring the earliest) scheduler 36.Each admittance in the dither controller 32 to 34 belongs to the input unit of the corresponding connection J1 to JM in the connection, keep them up to the corresponding virtual time of advent, and its information of ending together with them is sent to scheduler 36 thereupon, connect J1 to JM and pass node i and its next node (i+1).Refer to comprise the concluding time of the frame of this unit the virtual time of advent of Unit one, and refer to the concluding time that the time-delay of this unit of measuring from this virtual time of advent is placed restrictions on deadline.EDD scheduler 36 is the deadline of all unit in this scheduler relatively, and serves node (i+1) near that deadline of its deadline.According to the present invention, each dither controller is comprising the presynchronization frame counter of 32-1 or 34-1 for example; And should deadline information represent by the count value that this frame counter certainly comes.
For PSPS framing strategy, the right frame boundaries of each adjacent node should align so that the wobble information of upstream node (for example node i) can be reached downstream node (for example node (i+1)) effectively.This alignment procedure can be back to be moved during a call setting one connects.This destination node activates its frame counter and a count value is mail to adjacent upstream node from this frame counter.When receiving this value, this upstream node with the value of reflection relevant propagation delay as described below and other factor for this calling initialization corresponding to this frame counter that is connected.For each adjacent node to repeat always this process up to this source node integrated this right adjacent upstream node.
With reference to figure 4, show a sequential chart, be used for interpreted frame counter presynchronization process.Make S
I, jExpression connects a frame time started at the node i place of j.So keep following relation at two adjacent node i with (i+1):
s
I-1, j=s
I, j+ d
I, j+ θ
I+1.... equation (1) wherein, S
I+1, jThe frame time started that the node (i+1) of expression connection j is located; d
I, jExpression connects the time-delay at the node i place of j and places restrictions on; ι
I+1Propagation delay between expression node i and the node (i+1); And θ
I+1The corresponding non-coupling time-delay that expression is caused by the non-coupling in the boundary of time slot between node i and node (i+1).Set out by node (i+1) if carry a unit of frame counter information, when its frame counter value was Ci+1, then it will be at time S
I+1, j+ (K*T
j-C
I+1+ 1)+ι
I+1Arrive node i, the k round numbers, wherein, Tj represents to connect the frame sign of j, and hypothesis experience one time slot during a unit transmits.
Meaning this time of advent and counting the time quantum of having passed from the frame boundaries of node i is d
I, j+ 2 ι
I+1+ θ i+1-C
I+1+ 1, and thereby arrive next later time slot in this unit, be initialized to Ci behind the frame counter at node i place, its value is provided by following formula:
C
i=C
I+1-1-d
I, j-[2 τ
I+1+ θ
I+1] (McdTj) ... .. equation (2) finishes when this initial procedure, and frame counter counts down with the speed of time slot, and is reset when it reaches 0 value and is value Tj.In case frame counter such as above-mentioned having carried out synchronously in the call initiationc stage connects one, each frame counter can be worked independently.Below, for simplicity, propagation delay between node (i-1) and the node i and the additional delay sum ι that causes by the non-coupling in the boundary of time slot between node (i-1) and the node i
I 'Expression.
From another angle, frame counter is arrived the virtual time of advent at each node place so that can determine a unit, and this wobble information is not sent to another node by a node by presynchronization.More specifically, a frame counter Ci, j works under the condition that equation (3) is set up, frame counter Ci wherein, j is positioned at the node i place that connects j.
.... equation (3) is wherein
Represent that k unit arrives the virtual time of advent of the node i that connects j; Di, j represent to connect the time-delay at the node i place of j and place restrictions on; ι
I+1Propagation delay between expression node i and the node (i+1);
Expression connects the actual time of arrival of k the unit of node i of j; And
Expression
The time Counter Value.
As a result, the virtual time of advent of the node i of k unit arrival connection j
Can be by this k unit being arrived the actual time of arrival of the node i that connects j
With
The time the value C of frame counter
I, j
The phase Calais is calculated, because
Represented certainly
Extremely
Remaining time.
With reference to figure 5, show first embodiment of the presynchronization of the frame counter when placing restrictions on the frame sign that equals corresponding connection j in the time-delay at node place.White arrow is represented the unit that enters, and each all arrives node corresponding, and black arrow is represented the unit that leaves, and each all leaves node corresponding.If the time frame that connects j is at moment S
O, jBeginning, k unit arrives the virtual time of advent of a source node that connects j
Can be calculated as follows:
... equation (4) is wherein
The individual unit of expression (K-1) arrives the virtual time of advent of the source node that connects j; Tj represents to connect the frame sign of j;
Represent that K unit arrives the actual time of arrival of the source node that connects j; And
Represent an operator, this operator rounds its operand house tail and changes into smallest positive integral big or that equate with it than it.And can be calculated as follows the virtual time of advent of the node i of K unit arrival connection j:
... equation (5) is wherein
Represent that K unit arrives the virtual time of advent of the node i that connects j;
Represent that K unit arrives the virtual time of advent of the node (i-1) that connects j; Tj represents to connect the frame sign of j; And ι i represents the propagation delay between node (i-1) and the node i.When K unit arrived source node, it was kept its virtual time of advent
And
With
Between the time interval during be sent to next node 1.
Frame counter is assigned to each node of the connection j among Fig. 5.Each frame counter is deducted 1 and be reset the frame sign Tj that is set to connection j when its value becomes 0 according to above-mentioned presynchronization process presynchronization, each time slot.More specifically, frame counter be by consider between two adjacent nodes propagation delay and by presynchronization, so that Ci, j lags behind Ci-1, and the size of j is ι i ', Ci, j is a frame count that is positioned at the node i place that connects j, Ci-1, j are positioned at the frame counter that the node (i-1) that connects j is located, and ι i is the propagation delay between node (i-1) and the node i; Ci+1, j hysteresis Ci, the size of j is ι
I+1', Ci+1, j are positioned at the frame counter that the node (i+1) that connects j is located, and ι
I+1It is the propagation delay between node i and the node (i+1); And Ci+2, j hysteresis Ci+1, the size of j is ι
I+2', C
I+2, jBe to be positioned at the frame counter that the node (i+2) that connects j is located, and ι
I+2It is the propagation delay between the node (i+2) of node (i+1).
Suppose to work as Ci-1, the value of j is that 2 o'clock K unit arrive node (i-1), works as Ci-1, and the value of j is that 5 o'clock K unit leave to node i.Because the propagation delay between node (i-1) and the node i is 2, Ci, j hysteresis Ci-1, the size of j is 2.Therefore, work as Ci-1, the value of j is that Ci is worked as in K the unit that left node (i-1) at 5 o'clock, and the value of j still is 5 o'clock arrival unit i.Work as Ci then, the value of j is that 0 o'clock this K unit is considered to arrive node i, keeps 5 time slots, is sent to node (i+1) again.
Simultaneously, in a common framing strategy, the time-delay that is given in the first node place by the frame sign of corresponding connection is placed restrictions on, and this frame sign is an inverse of distributing to the bandwidth of this corresponding connection.Correspondingly, coupling is placed restrictions in the distribution of bandwidth and this time-delay, and this will cause the inefficacy of network resource usage.Can distribute from the distribution of bandwidth that decoupling zero is incompatible to alleviate this kind inefficacy by the boundary of will delaying time.
With reference to figure 6, second embodiment of the presynchronization of the frame counter when time-delay is placed restrictions on less than frame sign has been described, this example is with allocated bandwidth uncoupled example from time-delay is placed restrictions on.White arrow is represented the unit that enters, and each unit arrives a node, and black arrow is represented the unit that leaves, and each unit leaves a node.Has the virtual time of advent
Deadline of K unit at node i place of connection j become
But not
Wherein mean before this that in the deadline of K unit of node i K unit must be sent to the moment of a node (i+1), di, j are that the time-delay that connects the node i of j is placed restrictions on, and Tj is the frame sign that connects j.And, each frame counter that is positioned at each node that connects j is by presynchronization, arrive the node i that connects j so that when the value at the frame counter that is positioned at the node (i-1) that connects j is C, leave the unit of node (i-1), the value that is positioned at the frame counter of the node i that connects j with discovery is C-(Tj-di-1, j), wherein Tj is the frame sign that connects j, and di-1, j is that the time-delay that connects the node (i-1) of j is placed restrictions on.
One frame counter is assigned to each node of the connection j among Fig. 6.Each frame counter is deducted 1 according to above-mentioned presynchronization process by presynchronization, each time slot, and is reset to the size of this frame when its value becomes 0.More specifically, frame counter is by considering that the propagation delay between two adjacent nodes, the time-delay that connects each node of j place restrictions on, and connect the frame sign of j and by presynchronization, so that Ci, j lags behind Ci-1, and the size of j is ι i-(Tj-d
I-1, j), Ci, j is a frame counter that is positioned at the node i place that connects j, Ci-1, j is positioned at the frame counter that the node (i-1) that connects j is located, and ι i is the propagation delay between node (i-1) and the node i, and Tj is the frame sign that connects j, and di-1, j is that the time-delay that is positioned at node (i-1) that connects j is placed restrictions on; Ci+1, j hysteresis Ci, the size of j is ι
I+1-(T
j-d
I, j), Ci+1, j are positioned at the frame counter that the node (i+1) that connects j is located, ι
I+1Be the propagation delay between node i and the node (i+1), and di, j is that the time-delay that connects the node i of j is placed restrictions on; Ci+2, j hysteresis Ci+1, the size of j is ι
I+2-(T
j-d
I+1, j), Ci+2, j are frame counters that is positioned at the node (i+2) that connects j, ι
I+2Be the propagation delay between node (i+1) and the node (i+2), and di+1, j is that the time-delay that connects the node (i+1) of j is placed restrictions on.
Suppose to work as Ci-1, the value of j is that 2 o'clock K unit arrive node (i-1), and works as Ci-1, and the value of j is to leave to node i in 5 o'clock.It is 9 that the time-delay of node (i-1) is placed restrictions on, and the frame sign of connection j is 10.Because the propagation delay between node (i-1) and the node i is 3, Ci, j hysteresis Ci-1, the size of j is 3-(10-9)=2.Therefore, work as Ci-1, the value of j is to leave K unit of node (i-1) at 5 o'clock at Ci, and the value of j is 4 o'clock arrival node i.So K unit is considered at Ci, the value of j is 0 o'clock arrival node i, keeps 4 time slots, and is transferred into node (i+1).
With reference to figure 7, the 3rd embodiment of the presynchronization of frame counter when time-delay is placed restrictions on greater than frame sign has been described, this example is with allocated bandwidth uncoupled another example from time-delay is placed restrictions on.White arrow is represented the unit that enters, and each arrives a node, represents the unit that leaves and deceive arrow, and each leaves a node.By serving a plurality of unit in the integrated frame Tj, identical bandwidth allows bigger time-delay to place restrictions on, and the size of this integrated frame Tj ' is increased to n*Tj, and Tj is the primitive frame size that connects j, n is the integer greater than 1, and time-delay is placed restrictions on greater than (n-1) * T and is not less than n*Tj.In other words, n frame is integrated into a frame to serve n unit thus in this integrated image duration.Can calculate the virtual time of advent that arrives K unit of the node i that connects j by laxative remedy: maximum unit can be serviced during a frame period Tj ' in n the unit, as long as by using in equation (4)
Replace
In equation (5), use
Replace
To the presynchronization of frame counter can with second embodiment of the invention in the identical mode used carry out.Here it is, and frame counter is assigned to each node with the connection in 7, and the value of n is 2.Each frame counter is deducted and is rearranged into frame sign according to above-mentioned presynchronization process presynchronization, each time slot when its value becomes 0.More specifically, frame counter is by considering that the propagation delay between two adjacent nodes, the time-delay that connects each node of j place restrictions on, and connect the frame sign of j and by presynchronization, so that Ci, j lags behind Ci-1, the size of j be ι i-(Tj '-d
I-1, j), Ci, j is a frame counter that is positioned at the node i place that connects j, Ci-1, j is positioned at the frame counter that the node (i-1) that connects j is located, and ι i is the propagation delay between node (i-1) and the node i, and Tj is the frame sign that connects j, and di-1, j is that the time-delay that is positioned at node (i-1) that connects j is placed restrictions on; Ci+1, j hysteresis Ci, the size of j is ι
I+1-(T
j'-d
I, j), Ci-1, j are positioned at the frame counter that the node (i+1) that connects j is located, ι
I+1Be the propagation delay between node i and the node (i+1), and di, j is that the time-delay that connects the node i of j is placed restrictions on; Ci+2, j hysteresis Ci+1, the size of j is ι
I+2-(T
j'-d
I+1, j), Ci+2, j are frame counters that is positioned at the node (i+2) that connects j, ι
I+2Be the propagation delay between node (i+1) and the node (i+2), and di+1, j is that the time-delay that connects the node (i+1) of j is placed restrictions on.
Suppose to work as Ci-1, the value of j is that 8 o'clock K unit arrive node (i-1), and works as Ci-1, the value of j is that 4 o'clock (K+1) individual unit arrives node (i-1), works as Ci-1, and the value of j is that 14 o'clock K unit leave to node i, work as Ci-1, the value of j is that 11 o'clock (K+1) individual unit leave to node i.It is 16 that the time-delay of node (i-1) is placed restrictions on, and the integrated frame sign that connects j is 20.Because the propagation delay between node (i-1) and the node i is 4, and to connect that time-delay that the node (i-1) of j locates places restrictions on be 16, ci, and j lags behind ci-1, and the size of j is 4-(20-16)=0, and in other words, ci, j are synchronized with ci-1, j fully.Therefore, work as ci-1, the value of j is to leave K unit of node (i-1) at 14 o'clock at ci, and the value of j is 10 o'clock arrival node i, and works as ci-1, and the value of j is to leave (K+1) individual unit of node (i-1) at 6 o'clock at ci, and the value of j is 2 o'clock arrival node i.So K unit is considered at ci, the value of j is 0 o'clock arrival node i, keeps 10 time slots, and is sent to node (i+1); And (K+1) individual unit is considered to be in ci, and the value of j is 0 o'clock arrival node i, keeps 2 time slots, and is transferred into node (i+1).
According to three embodiment of the invention described above, can deduce a general algorithm be generalized to 0 (di, all possible time-delay in j＜∞ scope is placed restrictions on, di wherein, j is that a time-delay that connects the node i of j is placed restrictions on.A plurality of unit are serviced in a frame period, and the arbitrary value that is configured to less than this frame sign is placed restrictions in the time-delay of each node.So, the size of bandwidth no matter, all end-to-end time-delay demands can both be satisfied.In this general algorithm, be calculated as follows the virtual time of advent:
At a source node,
... .. equation (6) wherein,
It is the virtual time of advent that K unit arrives the source node that connects j; s
O, jBe to connect the time that the time frame at the source node place of j begins; N is in serviced unit number of an image duration;
Be the virtual time of advent that (k-n) individual unit arrives the source node that connects j, and when working as (k-n) and being less than or equal to " 0 ",
Be considered to s
O, jThough, there is no (k-n) individual unit; Tj is the frame sign that connects j;
Be that K unit arrives the source node actual time of arrival that connects j; And
Be an operator, it casts out the mantissa of operand so that it becomes the smallest positive integral more than or equal to this operand; And
At other node,
... equation (7) is wherein
It is the virtual time of advent that K unit arrives the node i that connects j; Di-1, j are that the time-delay that the node (i-1) of connection j is located is placed restrictions on; And ι i is the propagation delay between node (i-1) and the node i.
This general algorithm can be used to have in the network of constant propagation delay well, is considered to be in call establishment stage by synchronous frame counter because it depends on.Yet under some restriction, this algorithm also can be used to have in the network of variable propagation delay.Place restrictions on di-1 if connect the frame sign Tj of j greater than the time-delay that Tj locates more than or equal to the node (i-1) that connects j, j changes (ι i with corresponding propagation delay
Max-ι i
Min) sum, i.e. Tj 〉=d
I-1, j+ ι i
Max-ι i
Min, then can use this general algorithm.In this case, the definition of the virtual time of advent becomes
ι i ' (=ι i wherein
Max) maximum propagation time-delay between expression node (i-1) and the node i.
With reference to figure 8, show the 4th embodiment of the presynchronization of the frame counter that is used for variable propagation delay, wherein propagation delay is variable in preset range.Because can (for example seeing such as NTP (Network Time Protocol) by utilizing synchronously of the frame counter that the variation of propagation delay becomes complicated more, D.Mills ＂ Internet timesynchronization:the Network Time Protocol ＂, IEEE Trans.commun, vol.39, no.10 1482-1493 page or leaf, in October, 1991) conventional synchronization agreement is finished.
Suppose frame sign Tj=15, d is placed restrictions in the time-delay that the node (i-1) of connection j is located
I-1, j=10, the propagation delay between node (i-1) and the node i changes
And when being positioned at the frame counter Ci-1 that the node (i-1) that connects j is located, the value of j is to arrive K unit of node (i-1) at 2 o'clock at Ci-1, and the value of j is to leave node (i-1) at 10 o'clock.Because Tj is 15, and di-1, j is that 10, a K unit leaves node (i-1) and places restrictions on Zao 5 time slots than delaying time.So if the propagation delay between node (i-1) and the node i is minimum, then K unit is at the frame counter Ci that is positioned at the node i place that connects j, the value of j is 6 o'clock arrival node i; And if the propagation delay between node (i-1) and the node i is maximum, then K unit be at the frame counter Ci that is positioned at the node i place that connects j, and the value of j is to arrive node i at 5 o'clock.In Fig. 8, the minimum time-delay between K unit experience node (i-1) and the node i, so, work as Ci, the value of j is 6 o'clock K unit arrival node i.Then, K unit is considered to be in Ci, and the value of j is 0 o'clock arrival node i, keeps 6 time slots, and is sent to node (i+1).
According to 4 embodiment of the present invention that provide so far, the value of frame counter has two implications.For the unit in the dither controller, it means the time surplus of the virtual time of advent, and for the unit in the scheduler, it means the time surplus of deadline.For example, if unit than its deadline Zao 3 time slots serviced at a node place, this element arrives its next node to find that the value at the frame counter at this node place is 3, therefore, it enters scheduler be held 3 time slots in dither controller after.From the angle of scheduler, this unit seems to seem that it has experienced maximum delay.In other words, the moment that this element enters in the scheduler of present node is identical, and no matter formerly the node place obtains serving the time that is spent.
A unit is sent out as it serviced by last time slot that the node in this connection is in a frame.Thereby, handle all unit all as their by being served at interval in a regular time, size that wherein should the time interval is identical with the frame sign of this connection.
By utilizing plenary session words phase frame and adjusting frame sign and service time limit and place restrictions on corresponding biasing this frame counter synchronously by using with each time-delay, the present invention can overcome coupled problem intrinsic in the framing strategy, thereby allocation of network resources neatly.
Though the description of this invention can be made other change and variation only with reference to some preferred embodiment under the prerequisite that does not deviate from the spirit and scope of the present invention that propose in following claim.
Claims (15)
1, a kind of method of the network unit that is used for being ranked is wherein set up a connection that is used to call out from source node by a plurality of nodes to destination node, and at each node place one frame counter is set, and this method comprises the following steps:
(a) detect adjacent node between propagation delay and corresponding non-coupling time-delay, the frame sign of placing restrictions on and should be connected in the time-delay at each node place, wherein each adjacent node points to the downstream node of destination node to comprising a upstream node and that points to source node, and this corresponding non-coupling time-delay is what to be caused by the non-coupling in the boundary of time slot between a upstream node and the downstream node;
(b) according to adjacent node between propagation delay and corresponding non-coupling time-delay, the frame sign of placing restrictions on and should be connected in the time-delay at each node place, Active Frame counter also synchronously, each frame counter is positioned at each node place; And
(c) according to the value of respective frame counter, transmit this unit from source node to destination node.
2, the method for claim 1, wherein step (b) may further comprise the steps:
(b1) according to adjacent node between propagation delay place restrictions on this frame counter of initialization with the time-delay at each the node place that is connected at this;
(b2) each time slot is deducted 1 with the value of this frame counter, and the counter that its value becomes " 0 " reseted the size that is changed to this frame, wherein a frame that should connect is made up of a plurality of time slots, and this node of unit arrival and this unit leave a node and all finish in a time slot; And
(b3) repeating step (b2) arrives destination node up to this unit.
3, method as claimed in claim 2, wherein:
If the frame sign that is less than or equal to this connection is placed restrictions in all time-delays at each node place of this connection, then during the frame length of this connection, transmit this unit; And
If place restrictions on greater than (n-1) T and be not more than nT in the time-delay at a node place of this connection, wherein T is the frame length of this connection, then transmits n unit during nT, n be one greater than 1 integer.
4, method as claimed in claim 3, wherein:
If an adjacent node between propagation delay fix, then synchronous this frame counter, the unit that leaves a right upstream node of this adjacent node when the value that is positioned at the frame counter at upstream node place with box lunch is C arrives downstream node, the value that is positioned at the frame counter of this downstream node with discovery is C-(T-d), and wherein d places restrictions in the time-delay at this upstream node place; And
If an adjacent node between propagation delay in a preset range, be variable and the size of the frame of this connection more than or equal to placing restrictions in the time-delay at a right upstream node place of this adjacent node and poor sum between maximum propagation time-delay and minimum time-delay, then by using a conventional synchronization agreement to come this frame counter synchronously, above-mentioned maximum propagation time-delay is the maximum propagation time-delay between upstream node and its downstream node, and minimum time-delay is the minimum time-delay between upstream node and its downstream node.
5, method as claimed in claim 4, wherein when the value of the frame counter that is positioned at a node place becomes 0, Unit one by keeping this unit to serve in the virtual time of advent of this node until this unit, is sent to its downstream node with this unit at this node place quilt then in the time-delay at this node place is placed restrictions on.
6, method as claimed in claim 5, wherein this unit is defined as follows virtual time of advent of arriving the node place of this connection:
At source node,
VT wherein
O, j kIt is the virtual time of advent that K unit arrives the source node place that connects j; S
O, jIt is the frame time started that connects j; N is in serviced unit number of an image duration; VT
O, j K-nBe the virtual time of advent that (k-n) individual unit arrives the source node that connects j, and when (k-n) is less than or equal to " 0 ", VT
O, j K-nBe considered to S
O, jThough, do not have (k-n) individual unit; AT
O, jIt is the actual time of arrival that K unit arrives the source node that connects j;
Be an operator, it is cast out mantissa with an operand and gets smallest positive integral more than or equal to this operand, and j is the integer index of this connection, and K is the integer index of this unit; And
In node i,
VT
k i，j＝VT
k i-1，j+d
i-1，j+τ
i+θ
i
If the propagation delay between two adjacent nodes is the wherein VTi that fixes, j is the virtual time of advent of K unit to the node i that connects j; VTi-1, j are the virtual times of advent that K unit arrives the node (i-1) that connects j; Di-1, j place restrictions in the time-delay that the node (i-1) that connects j is located; ι i is the propagation delay between node (i-1) and the node i; And θ i represents the corresponding non-coupling time-delay between node (i-1) and the node i, and i is the integer index of this node, and j is the integer index of this connection, and K is the integer index of this unit; And
VT
k i，j＝VT
k i-1，j+d
i-1，j+τ
1 i+θ
i
If the propagation delay between two adjacent nodes is variable in a scope of being scheduled to,
Wherein ι i ' is the maximum propagation time-delay between node (i-1) and the node i.
7, method as claimed in claim 6, wherein calculate the virtual time of advent that this unit arrives the node i that connects j and carry out in the following manner: the actual time of arrival that this unit is arrived the node i that connects j adds the value of the frame counter when actual time of arrival node i place, this unit of adjunction j.
8, a kind of method is used for being lined up to Section Point from the first node of a network in a unit, and this first and second node has one first and one second counter respectively, and this method may further comprise the steps:
(a) place restrictions on all counters of presynchronization according to the time-delay of propagation delay between the node and first node, the value of each counter begins to count down from the value by a presynchronization program setting and deducts 1, up to 1, be reset to M then, and begin to count down from M and deduct 1, up to 1, M is one greater than 1 integer, and the value that is counted corresponding to P is placed restrictions in this time-delay, P is equal to or less than M, wherein by the presynchronization process presynchronization of all counters is carried out as follows: the value sum of second counter when count value of first counter is same as (M-P) with this arrival Section Point place, unit in this unit leaves first node;
(b) when this unit arrives first node, keep this unit to count down to 1 and be reset and be M up to first counter at this; And
(c) any sends to Section Point with this unit from first node before taking place in two kinds of situations below in the count value of first counter: if P less than M, then this count value reaches (M-P); If P equals M, then this value is reset to M.
9, method as claimed in claim 8, wherein each counter keeps a predetermined time interval with count value.
10, method as claimed in claim 9, wherein this unit leaves a node and this unit and arrives this node and all finish in this predetermined time interval.
11, method as claimed in claim 10, wherein this unit to leave and arrive all be that counter at this node place carries out when keeping a count value.
12, a kind of wherein these nodes comprise source node and destination node by the be ranked method of a plurality of unit of a calling of the connection that is made of a plurality of nodes, and each node of this connection has a frame counter, and this method may further comprise the steps:
(a) size with a frame is placed restrictions in a time-delay of determining each node of this calling, and wherein this calling is divided into a plurality of frames at each node place, and each frame comprises M time slot, and M is a positive integer, and each time slot is a time interval of presetting;
(b) determine a propagation delay between the two right nodes of each adjacent node, each adjacent node is to comprising a upstream node that points to source node and a downstream node that points to destination node;
(c) place restrictions on all frame counter presynchronization according to the time-delay of the right upstream node of the right propagation delay of each adjacent node and each adjacent node to this node, the time-delay of this upstream node is placed restrictions on corresponding to P time slot, P one is not more than the positive integer of M, the count down time slot of each frame of this node of the frame counter of one node, wherein the right counter of each adjacent node by presynchronization in case the count value of this upstream node that a unit is left equal (M-P) and this unit arrive this downstream node the time this downstream node the count value sum;
(d) when each unit arrives the right upstream node of the adjacent node that comprises source node, keep each unit to arrive the end of a frame up to this unit;
When (e) count value of the upstream node in step (d) satisfies one of following two conditions each unit is sent to its corresponding downstream node from this upstream node: if P less than M, then this count value reaches (M-P); If P equals M, then reseting this count value is M; And
(f) for each remaining adjacent node node that to repeating said steps (d) and (e) achieves the goal up to each unit.
13, method as claimed in claim 12, the wherein boundary alignment of the border of the time slot in a upstream node and the time slot in its corresponding downstream node.
14, method as claimed in claim 13, wherein said propagation delay detects step (b) and also comprises step (b1): detect the non-coupling time-delay between the right upstream and downstream node of each adjacent node, when this propagation delay amount greater than n Preset Time doubly at interval but be not more than (n+1) this Preset Time doubly at interval the time, this non-coupling time-delay is the absolute difference between long-pending corresponding to (n+1) that reset the time interval of the propagation delay amount of this upstream node and a time slot, and n is a positive integer.
15, method as claimed in claim 14, wherein the counter of the right upstream and downstream node of each adjacent node is by presynchronization, so that at the count value C of an instantaneous upstream node
iCount value C with downstream node
I+1Be defined as:
Wherein di represents that the time-delay of this upstream node places restrictions on; τ
I+1Represent the propagation delay between this upstream node and this downstream node; And θ
I+1Represent the non-coupling time-delay between this upstream node and this downstream node.
Applications Claiming Priority (12)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| KR49692/96 | 1996-10-29 | ||
| KR49692/1996 | 1996-10-29 | ||
| KR19960049692 | 1996-10-29 | ||
| KR60087/96 | 1996-11-29 | ||
| KR1019960060083A KR19980040845A (en) | 1996-11-29 | 1996-11-29 | Jitter information transfer method by counter interworking in asynchronous transmission mode communication network |
| KR60084/96 | 1996-11-29 | ||
| KR1019960060084A KR19980040846A (en) | 1996-11-29 | 1996-11-29 | Efficient Bandwidth Usage in Jitter Information Delivery by Counter Interlocking in Asynchronous Transfer Mode Network |
| KR1019960060087A KR19980040849A (en) | 1996-11-29 | 1996-11-29 | Improvement of Schedule Characteristics in Transfer of Jitter Information by Counter Interworking in Asynchronous Transmission Mode Networks |
| KR60087/1996 | 1996-11-29 | ||
| KR60083/96 | 1996-11-29 | ||
| KR60084/1996 | 1996-11-29 | ||
| KR60083/1996 | 1996-11-29 |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN1185072A CN1185072A (en) | 1998-06-17 |
| CN1115828C true CN1115828C (en) | 2003-07-23 |
Family
ID=27483164
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN97120230A Expired - Fee Related CN1115828C (en) | 1996-10-29 | 1997-10-29 | Pre-session pre-synchronized framing method for real-time services in ATM networks |
Country Status (5)
| Country | Link |
|---|---|
| US (1) | US6052389A (en) |
| JP (1) | JPH10145398A (en) |
| CN (1) | CN1115828C (en) |
| DE (1) | DE19747811B4 (en) |
| GB (1) | GB2321570B (en) |
Families Citing this family (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR100245329B1 (en) * | 1997-04-30 | 2000-02-15 | 전주범 | Apparatus of managing a queue by priority for controlling jitter in a packet network |
| US6381647B1 (en) * | 1998-09-28 | 2002-04-30 | Raytheon Company | Method and system for scheduling network communication |
| FR2872654B1 (en) * | 2004-06-30 | 2006-09-08 | Alcatel Sa | METHOD FOR RE-SYNCHRONIZING FLOWS BENEFITING FROM CIRCUIT EMULATION SERVICES FOR A PACKET SWITCHED COMMUNICATIONS NETWORK |
| KR100754736B1 (en) * | 2006-02-10 | 2007-09-03 | 삼성전자주식회사 | Method and apparatus for reproducing image frames in video receiver system |
| GB201717689D0 (en) * | 2017-10-27 | 2017-12-13 | Microsoft Technology Licensing Llc | Phase cashing for fast data recovery |
| CN113489328B (en) * | 2018-11-08 | 2024-02-27 | 立锜科技股份有限公司 | Zero-voltage switching control circuit, flyback power supply circuit and control method thereof |
| GB201905471D0 (en) | 2019-04-17 | 2019-05-29 | Microsoft Technology Licensing Llc | Amplitude caching in receive-from-many communications networks |
Family Cites Families (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5043982A (en) * | 1989-05-24 | 1991-08-27 | At&T Bell Laboratories | Multichannel multipoint network using time-division multiplexing incorporating a time offset for propagation delay |
| FR2669798B1 (en) * | 1990-11-23 | 1994-09-16 | Lmt Radio Professionelle | DEVICE FOR TRANSMITTING SYNCHRONOUS INFORMATION OVER AN ASYNCHRONOUS NETWORK, ESPECIALLY AN ATM NETWORK. |
| EP0639909A1 (en) * | 1993-08-17 | 1995-02-22 | ALCATEL BELL Naamloze Vennootschap | Resequencing system |
| GB9419802D0 (en) * | 1994-09-30 | 1994-11-16 | Roke Manor Research | Serial egress shaping for atm switch |
| US5570346A (en) * | 1994-12-08 | 1996-10-29 | Lucent Technologies Inc. | Packet network transit delay measurement system |
| US5648970A (en) * | 1996-03-04 | 1997-07-15 | Motorola, Inc. | Method and system for ordering out-of-sequence packets |
| JP3157718B2 (en) * | 1996-08-21 | 2001-04-16 | 日本電気エンジニアリング株式会社 | Time management method for ATM transmission network |
| US5933414A (en) * | 1996-10-29 | 1999-08-03 | International Business Machines Corporation | Method to control jitter in high-speed packet-switched networks |
-
1997
- 1997-10-24 JP JP9291914A patent/JPH10145398A/en active Pending
- 1997-10-27 GB GB9722688A patent/GB2321570B/en not_active Expired - Fee Related
- 1997-10-27 US US08/958,284 patent/US6052389A/en not_active Expired - Lifetime
- 1997-10-29 DE DE19747811A patent/DE19747811B4/en not_active Expired - Fee Related
- 1997-10-29 CN CN97120230A patent/CN1115828C/en not_active Expired - Fee Related
Also Published As
| Publication number | Publication date |
|---|---|
| GB2321570B (en) | 2001-01-24 |
| CN1185072A (en) | 1998-06-17 |
| DE19747811B4 (en) | 2007-05-16 |
| US6052389A (en) | 2000-04-18 |
| DE19747811A1 (en) | 1998-04-30 |
| GB9722688D0 (en) | 1997-12-24 |
| GB2321570A (en) | 1998-07-29 |
| JPH10145398A (en) | 1998-05-29 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| EP0591150B1 (en) | Congestion management based on multiple framing strategy | |
| KR100431191B1 (en) | An apparatus and method for scheduling packets by using a round robin based on credit | |
| US5408465A (en) | Flexible scheme for admission control of multimedia streams on integrated networks | |
| US6122673A (en) | Port scheduler and method for scheduling service providing guarantees, hierarchical rate limiting with/without overbooking capability | |
| US6724726B1 (en) | Method of putting a flow of packets of a network for transporting packets of variable length into conformity with a traffic contract | |
| EP1646192B1 (en) | Packet switch, scheduling device, drop control circuit, multicast control circuit and QOS control device | |
| JP2834293B2 (en) | How to change the virtual path capacity | |
| GB2324678A (en) | An earliest deadline first communications cell scheduler | |
| CN1115828C (en) | Pre-session pre-synchronized framing method for real-time services in ATM networks | |
| US5862127A (en) | Method of controlling the peak cell rate spacing of multiplexed ATM traffic | |
| Keshav et al. | Design and analysis of a flow control algorithm for a network of rate allocating servers | |
| Mowbray et al. | Capacity reservation for multimedia traffic | |
| Bensaou et al. | Credit-based fair queueing (CBFQ): A simple and feasible scheduling algorithm for packet networks | |
| Bondavalli et al. | MAC protocols for high-speed MANs: Performance comparisons for a family of Fasnet-based protocols | |
| Sathaye et al. | Conventional and early token release scheduling models for the IEEE 802.5 token ring | |
| Philp et al. | Scheduling and buffer management for soft-real-time VBR traffic in packet-switched networks | |
| Martinez et al. | Decoupling the bandwidth and latency bounding for table-based schedulers | |
| KR19980033218A (en) | Connection activation framing method for real-time service in ATM network | |
| Liebeherr et al. | Design and analysis of a high-performance packet multiplexer for multiservice networks with delay guarantees | |
| KR100221320B1 (en) | Method for determining the size of the per-session frame defined by the synchronus counter operation in atm networks | |
| KR100221317B1 (en) | Apparatus and method of the dynamic priority queueing dscipline using the per-session frame defined by the synchronus counter operation in atm networks | |
| KR100221316B1 (en) | A method for retrieving the jitter information for the virtual per-session frame by using a timer in atm networks | |
| KR0151911B1 (en) | Cell gap control apparatus and method for controlling through put band in atm network | |
| Comer et al. | Design and analysis of hybrid packet schedulers | |
| de Albuquerque et al. | Implementation of Policing Mechanisms for High Speed Networks |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| C41 | Transfer of patent application or patent right or utility model | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20031212 Address after: Seoul Patentee after: Daewoo Electronics Co., Ltd. Address before: Seoul Patentee before: Daewoo Electronics Co., Ltd. |
|
| ASS | Succession or assignment of patent right |
Owner name: FENGYE VISION TECHNOLOGY CO., LTD. Free format text: FORMER OWNER: DAEWOO ELECTRONICS CO., LTD. Effective date: 20130417 |
|
| C41 | Transfer of patent application or patent right or utility model | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20130417 Address after: Ontario Patentee after: Fengye Vision Technology Co., Ltd. Address before: Seoul, South Kerean Patentee before: Daewoo Electronics Co., Ltd. |
|
| CF01 | Termination of patent right due to non-payment of annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20030723 Termination date: 20161029 |